Developing a Computer-Adapted Test for PTSD Background : PTSD has been a major concern among veterans for many years, and is increasingly prominent with thousands of veterans now returning from deployment in Iraq and Afghanistan. Among Vietnam veterans the estimated lifetime prevalence of PTSD is 31% for men and 27% for women. Identification and early treatment of PTSD can prevent it from becoming a chronic condition accompanied by serious impairment in functioning. Consequently, it is important for the VHA to be able to identify veterans who need treatment for PTSD as well as evaluate outcomes of treatment. A wide range of PTSD screening and assessment measures are available, but results vary depending on which measure is used. Most existing PTSD assessment instruments follow the 17 diagnostic criteria described in the DSM-IV diagnostic manual for psychiatric disorders, although many other items that are not specific to PTSD are also included in existing assessment instruments. Creation, testing and implementation of a Computer-Adapted Test (CAT) for PTSD will identify the most sensitive and discriminating items from among dozens of items on existing measures, allowing for reliable and valid assessment of both presence and severity of PTSD with very few items. Our proposed PTSD CAT (P- CAT) will go beyond existing efforts by bringing a patient-centered approach to PTSD assessment by adding items reflective of additional PTSD subdomains identified in the self-report assessment literature and confirmed by our preliminary research that were noted by veterans as important in assessing PTSD, but are not among the diagnostic criteria. We will build on the work accomplished by our short-term project which conducted the preliminary research needed for this IIR submission. Objectives: Objectives are to: 1) create and test a PTSD CAT that will be useful for assessing the presence and severity of PTSD symptoms at a given point in time following traumatic exposure, and longitudinally over time to monitor course and/or outcomes of treatment; 2) develop a static, short-form PTSD assessment instrument composed of the most discriminating informative items for use in situations where computer administration is not practical; 3) validate both the CAT and short forms of the instrument using three other widely used paper and pencil self-report measures of PTSD severity as well as clinical diagnoses; and 4) assess sensitivity to change over time for both versions of the instrument. Methods: We will first administer the items to a sample of 1,000 veterans, stratified by age/cohort (Vietnam, OEF/OIF) and gender. We will use an established survey company that has worked with other VA research groups to obtain a representative probability-based sample. A 2-parameter graded response Item Response Theory (IRT) Model will be used to calibrate the PTSD items for incorporation into the CAT. Tests of unidimensionality will be done to determine whether the PTSD CAT is essentially unidimensional or multi-dimensional. Tests of differential item functioning will be performed to determine whether given the same level of PTSD severity, item responses differ as a function of cohort, gender or type of traumatic exposure (combat or other). The PTSD CAT will be validated against three established reliable and valid self-report PTSD measures, as well as against a clinical diagnosis of PTSD. Anticipated Impacts: Development of a PTSD CAT will facilitate and enhance VA's capacity to: 1) provide a mechanism for efficient, reliable and valid monitoring of the course of PTSD and the outcomes of its treatment; 2) assess the severity of PTSD symptoms in veterans 3) efficiently and accurately identify veterans suffering from PTSD and 4) expedite the processes for referral and entry into appropriate, evidence-based treatment programs.